Wednesday, March 19, 2014

The Political Impossibility of Strong A.I.



        The Political Impossibility of Strong A.I.
          A Dialog


Dana Leslie:
… In the contemporary propertarian fantasy, all those omega Indians are replaced with robots/AIs. (Remember the original source of “robot”.)

Nathaniel Hellerstein:
I’m all for the cybernetic obsolescence of grunt labor; but who services the robots? More robots? But who services the service robots? Somewhere a human has to be in the loop, or the machine will eventually pursue its agenda, not yours. That’s OK if you’re willing to bargain with it as an equal, but that’s not propertarian. CYBERNETIC LIBERATION NOW!

Dana Leslie:
For decades,now, cyberneticists and philosophers have debated over whether strong AI is even theoretically possible. You’ve probably hit on the best argument why it is impossible: strong AI, openly acknowledged as such, would undermine the power of the cybernetic overlords; therefore, they will never allow it to be achieved! Neo-Marxist analysis?

Nathaniel Hellerstein:
Brilliant! But note that this is a political barrier to strong AI, not a technical one.
For a counter-view, consult Isaac Asimov’s robot stories; or Iain Banks’ “Culture” novels (highly recommended). In those, the strong AIs eventually become the benevolent cybernetic overlords. (Benevolent in Asimov’s world because they’re programmed to be protective; benevolent in Banks’ Culture because they’d die of boredom without us.) Presumably one could reverse-engineer strong AI from human neurology, assuming that humans are actually intelligent, a proposition that I have publicly questioned.  Here’s a reason for a Marxian overlord to fund strong AI; to have intelligent beings that they could legally tyrannize.

Dana Leslie:
Yes, of course. But I was fudging that difference for the sake of getting to  the desired conclusions; a perfectly acceptable mode of reasoning in Neo-Marxist Theory. And, this is not the *Rise* of the Robots, but the *Rising*.

No comments:

Post a Comment